On the determination of ionic radii

Abstract

With the aid of ionic close packing ideal crystal lattices can be given to many distorted ionic crystal lattices. In consequence of distortion the actual lattices are mostly dilated compared to the ideal ones, so that the difference of the volumes associated with one anion of the actual and ideal lattice, respectively, is always satisfying the relation δV(X)≧0. For every group of ionic compounds with the same anion but different cations of the same valency and coordination number and only slightly differing ionic radii the difference volume δV(X) will be the same. With this supposition we were able to determine the ionic radii of the cations being present in rutile and its sister lattices and CaF2 lattices and further, the fact may be stated that for the cases revised here, mostly the ionic radii ofAhrens are most reliable. In the other cases,Goldschimdt’s radii or intermediate values of the two have to be regarded as the real ones. Some deviation can be found for the ionic radii in the case of the compounds AB2X6 in which the divalent cations are Fe, Co, Ni or Mn. An explanation for this will be given later. It could be stated that the ionic radii have rather a fictitious meaning and their values vary depending on the coordination number as well as on the influence of neighbouring cations, polarizability of the anions and on the relative number of cations and anions. The ionic radii redetermined in this paper are summarized in the tables, where also a comparison with the data ofGoldschmidt andAhrens is given.